Conference Object

Response rates in psychological online surveys. A meta-analysis on the effects of study design and time

Author(s) / Creator(s)

Burgard, Tanja
Kasten, Nadine
Bosnjak, Michael

Abstract / Description

Relevance & Research Question: The main question of the meta-analysis is, whether the initial participation rate in psychological studies has decreased over time. Moreover, possible moderators of this time effect will be addressed: The design of an invitation letter, the contact protocol, the data collection mode, the burden of participating in the study and the incentives given to participants. Methods & Data: Eligible studies for the meta-analysis have to report (quasi-)experiments on initial response rates from empirical studies in the field of psychology. The experimental manipulation of an eligible study is the variation of survey design characteristics. Student samples will be excluded, because students are often obliged to participate for their studies and therefore, their motivation differs from other populations. The outcome of interest will be the initial response rate. As there may be different experimental comparisons per study report, the data are hierarchical. Using the metafor package in R, three-level mixed effects models will be used to account for the dependencies in the data and to enable testing moderator variables on the level of the report (e.g. Type of report, publication year) and on the level of the experiment (e.g. year of data collection, incentives). The relevant independent variable for all tests is the time of sampling. The moderating effects of the survey design will be tested using the characteristics of study conduction as moderator variables. Results: Results are not available yet. Added Value: The trend of declining response rates in the last decades can aggravate the possible bias due to nonresponse. Therefore, it is of interest what factors may moderate this trend to be able to guide survey operations by empirical evidence to optimize survey response. Due to the change in the willingness to participate in scientific studies, the continuous updating of the evidence is of importance.

Persistent Identifier

Date of first publication

2019-07-16

Is part of

Conference of the European Survey Research Association (ESRA) 2019, Zagreb, Croatia

Publisher

PsychArchives

Citation

Burgard, T., Kasten, N., & Bosnjak, M. (2019, July 16). Response rates in psychological online surveys. A meta-analysis on the effects of study design and time. PsychArchives. https://doi.org/10.23668/psycharchives.2515
  • Author(s) / Creator(s)
    Burgard, Tanja
  • Author(s) / Creator(s)
    Kasten, Nadine
  • Author(s) / Creator(s)
    Bosnjak, Michael
  • PsychArchives acquisition timestamp
    2019-07-17T13:22:57Z
  • Made available on
    2019-07-17T13:22:57Z
  • Date of first publication
    2019-07-16
  • Abstract / Description
    Relevance & Research Question: The main question of the meta-analysis is, whether the initial participation rate in psychological studies has decreased over time. Moreover, possible moderators of this time effect will be addressed: The design of an invitation letter, the contact protocol, the data collection mode, the burden of participating in the study and the incentives given to participants. Methods & Data: Eligible studies for the meta-analysis have to report (quasi-)experiments on initial response rates from empirical studies in the field of psychology. The experimental manipulation of an eligible study is the variation of survey design characteristics. Student samples will be excluded, because students are often obliged to participate for their studies and therefore, their motivation differs from other populations. The outcome of interest will be the initial response rate. As there may be different experimental comparisons per study report, the data are hierarchical. Using the metafor package in R, three-level mixed effects models will be used to account for the dependencies in the data and to enable testing moderator variables on the level of the report (e.g. Type of report, publication year) and on the level of the experiment (e.g. year of data collection, incentives). The relevant independent variable for all tests is the time of sampling. The moderating effects of the survey design will be tested using the characteristics of study conduction as moderator variables. Results: Results are not available yet. Added Value: The trend of declining response rates in the last decades can aggravate the possible bias due to nonresponse. Therefore, it is of interest what factors may moderate this trend to be able to guide survey operations by empirical evidence to optimize survey response. Due to the change in the willingness to participate in scientific studies, the continuous updating of the evidence is of importance.
    en_US
  • Citation
    Burgard, T., Kasten, N., & Bosnjak, M. (2019, July 16). Response rates in psychological online surveys. A meta-analysis on the effects of study design and time. PsychArchives. https://doi.org/10.23668/psycharchives.2515
    en
  • Persistent Identifier
    https://hdl.handle.net/20.500.12034/2139
  • Persistent Identifier
    https://doi.org/10.23668/psycharchives.2515
  • Language of content
    eng
    en_US
  • Publisher
    PsychArchives
    en_US
  • Is part of
    Conference of the European Survey Research Association (ESRA) 2019, Zagreb, Croatia
    en_US
  • Dewey Decimal Classification number(s)
    150
  • Title
    Response rates in psychological online surveys. A meta-analysis on the effects of study design and time
    en_US
  • DRO type
    conferenceObject
    en_US
  • Leibniz institute name(s) / abbreviation(s)
    ZPID